Action delays degrade the performance of reinforcement learning in many real-world systems. This paper proposes a formal definition of delay-aware Markov Decision Process and proves it can be transformed into standard MDP with augmented states using the Markov reward process. We develop a delay-aware model-based reinforcement learning framework that can incorporate the multi-step delay into the learned system models without learning effort. Experiments with the Gym and MuJoCo platforms show that the proposed delay-aware model-based algorithm is more efficient in training and transferable between systems with various durations of delay compared with off-policy model-free reinforcement learning methods. Codes available at: https://github.com/baimingc/dambrl.
Reinforcement learning (RL) enables robots to learn skills from interactions with the real world. In practice, the unstructured step-based exploration used in Deep RL -- often very successful in simulation -- leads to jerky motion patterns on real robots. Consequences of the resulting shaky behavior are poor exploration, or even damage to the robot. We address these issues by adapting state-dependent exploration (SDE) to current Deep RL algorithms. To enable this adaptation, we propose two extensions to the original SDE, using more general features and re-sampling the noise periodically, which leads to a new exploration method generalized state-dependent exploration (gSDE). We evaluate gSDE both in simulation, on PyBullet continuous control tasks, and directly on three different real robots: a tendon-driven elastic robot, a quadruped and an RC car. The noise sampling interval of gSDE permits to have a compromise between performance and smoothness, which allows training directly on the real robots without loss of performance. The code is available at https://github.com/DLR-RM/stable-baselines3.
The overestimation bias is one of the major impediments to accurate off-policy learning. This paper investigates a novel way to alleviate the overestimation bias in a continuous control setting. Our method---Truncated Quantile Critics, TQC,---blends three ideas: distributional representation of a critic, truncation of critics prediction, and ensembling of multiple critics. Distributional representation and truncation allow for arbitrary granular overestimation control, while ensembling provides additional score improvements. TQC outperforms the current state of the art on all environments from the continuous control benchmark suite, demonstrating 25% improvement on the most challenging Humanoid environment.
Adversarial Imitation Learning (AIL) is a class of algorithms in Reinforcement learning (RL), which tries to imitate an expert without taking any reward from the environment and does not provide expert behavior directly to the policy training. Rather, an agent learns a policy distribution that minimizes the difference from expert behavior in an adversarial setting. Adversarial Inverse Reinforcement Learning (AIRL) leverages the idea of AIL, integrates a reward function approximation along with learning the policy, and shows the utility of IRL in the transfer learning setting. But the reward function approximator that enables transfer learning does not perform well in imitation tasks. We propose an Off-Policy Adversarial Inverse Reinforcement Learning (Off-policy-AIRL) algorithm which is sample efficient as well as gives good imitation performance compared to the state-of-the-art AIL algorithm in the continuous control tasks. For the same reward function approximator, we show the utility of learning our algorithm over AIL by using the learned reward function to retrain the policy over a task under significant variation where expert demonstrations are absent.
In this paper, we present a new reinforcement learning (RL) algorithm called Distributional Soft Actor Critic (DSAC), which exploits the distributional information of accumulated rewards to achieve better performance. Seamlessly integrating SAC (which uses entropy to encourage exploration) with a principled distributional view of the underlying objective, DSAC takes into consideration the randomness in both action and rewards, and beats the state-of-the-art baselines in several continuous control benchmarks. Moreover, with the distributional information of rewards, we propose a unified framework for risk-sensitive learning, one that goes beyond maximizing only expected accumulated rewards. Under this framework we discuss three specific risk-related metrics: percentile, mean-variance and distorted expectation. Our extensive experiments demonstrate that with distribution modeling in RL, the agent performs better for both risk-averse and risk-seeking control tasks.
Deterministic-policy actor-critic algorithms for continuous control improve the actor by plugging its actions into the critic and ascending the action-value gradient, which is obtained by chaining the actor's Jacobian matrix with the gradient of the critic with respect to input actions. However, instead of gradients, the critic is, typically, only trained to accurately predict expected returns, which, on their own, are useless for policy optimization. In this paper, we propose MAGE, a model-based actor-critic algorithm, grounded in the theory of policy gradients, which explicitly learns the action-value gradient. MAGE backpropagates through the learned dynamics to compute gradient targets in temporal difference learning, leading to a critic tailored for policy improvement. On a set of MuJoCo continuous-control tasks, we demonstrate the efficiency of the algorithm in comparison to model-free and model-based state-of-the-art baselines.
We propose a simple data augmentation technique that can be applied to standard model-free reinforcement learning algorithms, enabling robust learning directly from pixels without the need for auxiliary losses or pre-training. The approach leverages input perturbations commonly used in computer vision tasks to regularize the value function. Existing model-free approaches, such as Soft Actor-Critic (SAC), are not able to train deep networks effectively from image pixels. However, the addition of our augmentation method dramatically improves SAC's performance, enabling it to reach state-of-the-art performance on the DeepMind control suite, surpassing model-based (Dreamer, PlaNet, and SLAC) methods and recently proposed contrastive learning (CURL). Our approach can be combined with any model-free reinforcement learning algorithm, requiring only minor modifications. An implementation can be found at https://sites.google.com/view/data-regularized-q.
We are interested in derivative-free optimization of high-dimensional functions. The sample complexity of existing methods is high and depends on problem dimensionality, unlike the dimensionality-independent rates of first-order methods. The recent success of deep learning suggests that many datasets lie on low-dimensional manifolds that can be represented by deep nonlinear models. We therefore consider derivative-free optimization of a high-dimensional function that lies on a latent low-dimensional manifold. We develop an online learning approach that learns this manifold while performing the optimization. In other words, we jointly learn the manifold and optimize the function. Our analysis suggests that the presented method significantly reduces sample complexity. We empirically evaluate the method on continuous optimization benchmarks and high-dimensional continuous control problems. Our method achieves significantly lower sample complexity than Augmented Random Search, Bayesian optimization, covariance matrix adaptation (CMA-ES), and other derivative-free optimization algorithms.
Standard planners for sequential decision making (including Monte Carlo planning, tree search, dynamic programming, etc.) are constrained by an implicit sequential planning assumption: The order in which a plan is constructed is the same in which it is executed. We consider alternatives to this assumption for the class of goal-directed Reinforcement Learning (RL) problems. Instead of an environment transition model, we assume an imperfect, goal-directed policy. This low-level policy can be improved by a plan, consisting of an appropriate sequence of sub-goals that guide it from the start to the goal state. We propose a planning algorithm, Divide-and-Conquer Monte Carlo Tree Search (DC-MCTS), for approximating the optimal plan by means of proposing intermediate sub-goals which hierarchically partition the initial tasks into simpler ones that are then solved independently and recursively. The algorithm critically makes use of a learned sub-goal proposal for finding appropriate partitions trees of new tasks based on prior experience. Different strategies for learning sub-goal proposals give rise to different planning strategies that strictly generalize sequential planning. We show that this algorithmic flexibility over planning order leads to improved results in navigation tasks in grid-worlds as well as in challenging continuous control environments.
The exploration-exploitation trade-off is at the heart of reinforcement learning (RL). However, most continuous control benchmarks used in recent RL research only require local exploration. This led to the development of algorithms that have basic exploration capabilities, and behave poorly in benchmarks that require more versatile exploration. For instance, as demonstrated in our empirical study, state-of-the-art RL algorithms such as DDPG and TD3 are unable to steer a point mass in even small 2D mazes. In this paper, we propose a new algorithm called "Plan, Backplay, Chain Skills" (PBCS) that combines motion planning and reinforcement learning to solve hard exploration environments. In a first phase, a motion planning algorithm is used to find a single good trajectory, then an RL algorithm is trained using a curriculum derived from the trajectory, by combining a variant of the Backplay algorithm and skill chaining. We show that this method outperforms state-of-the-art RL algorithms in 2D maze environments of various sizes, and is able to improve on the trajectory obtained by the motion planning phase.
We present differentiable predictive control (DPC), a method for learning constrained neural control policies for linear systems with probabilistic performance guarantees. We employ automatic differentiation to obtain direct policy gradients by backpropagating the model predictive control (MPC) loss function and constraints penalties through a differentiable closed-loop system dynamics model. We demonstrate that the proposed method can learn parametric constrained control policies to stabilize systems with unstable dynamics, track time-varying references, and satisfy nonlinear state and input constraints. In contrast with imitation learning-based approaches, our method does not depend on a supervisory controller. Most importantly, we demonstrate that, without losing performance, our method is scalable and computationally more efficient than implicit, explicit, and approximate MPC. Under review at IEEE Transactions on Automatic Control.
In this paper, we propose a multi-timescale replay (MTR) buffer for improving continual learning in RL agents faced with environments that are changing continuously over time at timescales that are unknown to the agent. The basic MTR buffer comprises a cascade of sub-buffers that accumulate experiences at different timescales, enabling the agent to improve the trade-off between adaptation to new data and retention of old knowledge. We also combine the MTR framework with invariant risk minimization, with the idea of encouraging the agent to learn a policy that is robust across the various environments it encounters over time. The MTR methods are evaluated in three different continual learning settings on two continuous control tasks and, in many cases, show improvement over the baselines.
We present CURL: Contrastive Unsupervised Representations for Reinforcement Learning. CURL extracts high-level features from raw pixels using contrastive learning and performs off-policy control on top of the extracted features. CURL outperforms prior pixel-based methods, both model-based and model-free, on complex tasks in the DeepMind Control Suite and Atari Games showing 1.9x and 1.2x performance gains at the 100K environment and interaction steps benchmarks respectively. On the DeepMind Control Suite, CURL is the first image-based algorithm to nearly match the sample-efficiency of methods that use state-based features. Our code is open-sourced and available at https://github.com/MishaLaskin/curl.
Potential Based Reward Shaping combined with a potential function based on appropriately defined abstract knowledge has been shown to significantly improve learning speed in Reinforcement Learning. MultiGrid Reinforcement Learning (MRL) has further shown that such abstract knowledge in the form of a potential function can be learned almost solely from agent interaction with the environment. However, we show that MRL faces the problem of not extending well to work with Deep Learning. In this paper we extend and improve MRL to take advantage of modern Deep Learning algorithms such as Deep Q-Networks (DQN). We show that DQN augmented with our approach perform significantly better on continuous control tasks than its Vanilla counterpart and DQN augmented with MRL.
Intrinsic motivation enables reinforcement learning (RL) agents to explore when rewards are very sparse, where traditional exploration heuristics such as Boltzmann or e-greedy would typically fail. However, intrinsic exploration is generally handled in an ad-hoc manner, where exploration is not treated as a core objective of the learning process; this weak formulation leads to sub-optimal exploration performance. To overcome this problem, we propose a framework based on multi-objective RL where both exploration and exploitation are being optimized as separate objectives. This formulation brings the balance between exploration and exploitation at a policy level, resulting in advantages over traditional methods. This also allows for controlling exploration while learning, at no extra cost. Such strategies achieve a degree of control over agent exploration that was previously unattainable with classic or intrinsic rewards. We demonstrate scalability to continuous state-action spaces by presenting a method (EMU-Q) based on our framework, guiding exploration towards regions of higher value-function uncertainty. EMU-Q is experimentally shown to outperform classic exploration techniques and other intrinsic RL methods on a continuous control benchmark and on a robotic manipulator.
Reinforcement learning (RL) is a powerful framework for learning to take actions to solve tasks. However, in many settings, an agent must winnow down the inconceivably large space of all possible tasks to the single task that it is currently being asked to solve. Can we instead constrain the space of tasks to those that are semantically meaningful? In this work, we introduce a framework for using weak supervision to automatically disentangle this semantically meaningful subspace of tasks from the enormous space of nonsensical "chaff" tasks. We show that this learned subspace enables efficient exploration and provides a representation that captures distance between states. On a variety of challenging, vision-based continuous control problems, our approach leads to substantial performance gains, particularly as the complexity of the environment grows.
Parameter space exploration methods with black-box optimization have recently been shown to outperform state-of-the-art approaches in continuous control reinforcement learning domains. In this paper, we examine reasons why these methods work better and the situations in which they are worse than traditional action space exploration methods. Through a simple theoretical analysis, we show that when the parametric complexity required to solve the reinforcement learning problem is greater than the product of action space dimensionality and horizon length, exploration in action space is preferred. This is also shown empirically by comparing simple exploration methods on several toy problems.
Model-free deep reinforcement learning (RL) has demonstrated its superiority on many complex sequential decision-making problems. However, heavy dependence on dense rewards and high sample-complexity impedes the wide adoption of these methods in real-world scenarios. On the other hand, imitation learning (IL) learns effectively in sparse-rewarded tasks by leveraging the existing expert demonstrations. In practice, collecting a sufficient amount of expert demonstrations can be prohibitively expensive, and the quality of demonstrations typically limits the performance of the learning policy. In this work, we propose Self-Adaptive Imitation Learning (SAIL) that can achieve (near) optimal performance given only a limited number of sub-optimal demonstrations for highly challenging sparse reward tasks. SAIL bridges the advantages of IL and RL to reduce the sample complexity substantially, by effectively exploiting sup-optimal demonstrations and efficiently exploring the environment to surpass the demonstrated performance. Extensive empirical results show that not only does SAIL significantly improve the sample-efficiency but also leads to much better final performance across different continuous control tasks, comparing to the state-of-the-art.
This paper introduces the Attribute-Decomposed GAN, a novel generative model for controllable person image synthesis, which can produce realistic person images with desired human attributes (e.g., pose, head, upper clothes and pants) provided in various source inputs. The core idea of the proposed model is to embed human attributes into the latent space as independent codes and thus achieve flexible and continuous control of attributes via mixing and interpolation operations in explicit style representations. Specifically, a new architecture consisting of two encoding pathways with style block connections is proposed to decompose the original hard mapping into multiple more accessible subtasks. In source pathway, we further extract component layouts with an off-the-shelf human parser and feed them into a shared global texture encoder for decomposed latent codes. This strategy allows for the synthesis of more realistic output images and automatic separation of un-annotated attributes. Experimental results demonstrate the proposed method's superiority over the state of the art in pose transfer and its effectiveness in the brand-new task of component attribute transfer.
Reinforcement learning (RL) has proven its worth in a series of artificial domains, and is beginning to show some successes in real-world scenarios. However, much of the research advances in RL are hard to leverage in real-world systems due to a series of assumptions that are rarely satisfied in practice. In this work, we identify and formalize a series of independent challenges that embody the difficulties that must be addressed for RL to be commonly deployed in real-world systems. For each challenge, we define it formally in the context of a Markov Decision Process, analyze the effects of the challenge on state-of-the-art learning algorithms, and present some existing attempts at tackling it. We believe that an approach that addresses our set of proposed challenges would be readily deployable in a large number of real world problems. Our proposed challenges are implemented in a suite of continuous control environments called the realworldrl-suite which we propose an as an open-source benchmark.